JMIR Publications

ABSTRACT

Artificial intelligence (AI), particularly large-scale language models (LLMs) such as ChatGPT, has emerged as a technology of significant impact in various fields, including medicine. This rapid development presents both opportunities and risks, particularly in the context of emergency medicine, where AI could transform clinical practices, but also raises concerns regarding the safety and reliability of its applications. This update aims to evaluate the implications of AI in the medical field, examining its potential applications in emergency medicine, its benefits and limitations, and the challenges of achieving general artificial intelligence (GAI). A literature review was conducted to analyze the current capabilities of AIs in health data processing, medical imaging, and clinical process improvement, while addressing concerns raised by hallucination phenomena and LLM performance in the context of rare or atypical cases. AI models offer substantial advantages in triage, patient flow optimization, bed management, and care prioritization in emergency medicine. However, significant risks remain, including AI hallucinations that can generate erroneous information and LLM limitations for infrequent clinical situations, potentially compromising patient safety. AI represents revolutionary potential for emergency medicine, but it necessitates a rigorous regulatory and safety approach to mitigate associated risks. The implementation of safety standards and supervisory practices becomes essential to ensure the safe and effective integration of AI into clinical medicine.